The Centers for Disease Control and Prevention seem more scared than elated by United States smoking rates. They have migrated from a war on the world's top killer, smoking, to being in a war on a chemical, nicotine.

They needn't be concerned. Science and health have won, and it wasn't because of taxes on cigarettes or a cottage industry of anti-smoking ads built by a tobacco company settlement, it was because of peer pressure. In young people. As the American Council on Science and Health, a pro-science consumer advocacy group based in New York City, has said since the 1970s, smoking is a pediatric disease. In the past, 90 percent of smokers picked up the habit by age 18, making adolescence a critical time for smoking-prevention efforts.

Peer influence has long been known as a major risk factor for adolescent smoking, but findings have varied about how big the risk is or how this dynamic unfolds. A meta-analysis of 75 longitudinal teen smoking studies published in Psychological Bulletin finds that having friends who smoke doubles the risk that children ages 10 to 19 will start smoking and continue smoking.

Less surprisingly, peer influence is more powerful in collectivist cultures, like Iceland or China, than in those where individualism is the norm, like America.

"We predicted that people in collectivistic cultures would be more likely to be influenced by peers around them, and that held true," says the study's lead author, Jiaying Liu, PhD, a recent graduate of the Annenberg School for Communication at the University of Pennsylvania. "In those most collectivistic countries, adolescents whose peers smoke are 4.3 times more likely to pick up smoking compared to people who have no peers smoking. By contrast, having peers who smoke made adolescents from those most individualist countries 1.89 times more likely to smoke."

If peer pressure matters, the opposite it also true. Those socialist or communist countries should also have less smoking. Except countries like China do not. The world's leading cigarette company is Chinese. Instead, peer pressure is simply different, not more profound.

The analysis used data from 16 countries, both collectivistic -- for example, China, South Korea, Jordan and Portugal -- and individualistic like the United States, Australia, Canada, the Netherlands and the United Kingdom.

The authors also examined the ethnic origins of the adolescents in each study, regardless of nationality. "We found that peer influence was much weaker in samples with higher proportions of adolescents with a European background, but much stronger in samples with higher proportions of adolescents with an Asian background," Liu says.

This finding about ethnic and cultural background may also explain why the results of teen smoking studies have varied in how big an effect peer influence has.

In addition, the study found that closer friends were more likely to influence peers to start smoking when compared to more distant friends. The closeness of peer friendships did not impact whether adolescents who smoke continued to smoke, however, indicating that perhaps the addictive nature of tobacco was an overriding factor.

The study was also able to better avoid what is often seen as the "chicken or the egg" problem of peer influence smoking studies: Do teens influence one another to smoke, or do those who smoke simply tend to become friends?

"By including only longitudinal studies, where peer influence was measured at an earlier time and adolescent smoking outcomes at a later date, we were better able to establish that peer influence led to the adolescent smoking outcomes, rather than the other way around," says Liu, who will be starting in the fall as an assistant professor in the University of Georgia's Department of Communication Studies.

The researchers hope that by providing a better understanding of the risk factors for teen smoking, they can help inform more targeted prevention efforts. By knowing who is most at risk, parents and public health officials alike can give teenagers the tools to resist tobacco.